PROJECT PROPOSAL

VERSION: Revision 1
Draft

Automated Troubleshooting, Maintenance and Diagnostics

For Internal Technical Support – an Expert System Approach

Author: Shubhadha Iyer

Last Revision: September 01,2000
 TABLE OF CONTENTS

The Future of Technical Troubleshooting and Support Solutions..............................................................3

Proposal for the development of Automated Troubleshooting solutions – an Expert System Approach.. .4
Overview:.....................................................................................................................................................4
Salient Features...........................................................................................................................................6
System Design..................................................................................................................................................7
Market Analysis...............................................................................................................................................8
Target Markets.................................................................................................................................................9
Operating Plan..................................................................................................................................................9
Sales Forecast...................................................................................................................................................9
Projections......................................................................................................................................................10
CONFIDENTIAL Page 3 5/10/2010

The Future of Technical Troubleshooting and Support Solutions

_________________________________________________________________

Through its revolutionary and groundbreaking Application Service solutions, XYZ

1
Systems Inc. is bringing a new and innovative approach to Troubleshooting and
Internal Technical Support.

This project aims to reduce costs associated with problem resolution and increase
productivity in the Call Center environment. XYZ Systems Inc., a privately held
company located in California and New York, prides itself on being able to remain
focused on the new challenges in diagnostics and maintenance that test IT
personnel and their troubleshooting capabilities.

The project is led by a team of experienced programmers and designers whose

background includes working with Knowledge-based Diagnostic Expert Systems.
In collaboration with a small group of talented and experienced associates, they are
preparing to offer a tool based on Artificial Intelligence technology that will
provide innovative and efficient solutions in this underdeveloped area.

Our vision for the future is designed with advanced features such as an interactive
diagnostic aid that eliminates the need for IT personnel to use cumbersome
troubleshooting manuals. The result is a productive IT technician who can spend
less time on repetitive maintenance and analysis, and thus contribute much more to
the overall profitability of their company.

Working closely with your company’s technical staff, the project team will closely
supervise and implement a system that is compatible with the existing architecture
through a smooth and uninterrupted transition. XYZ Systems Inc1. has the facilities
and technical capabilities to ensure the proper operation of its systems in any
setting. The applications will go through a stringent testing process in order to
verify their complete reliability.

Meeting the challenges of the future and providing breakthrough solutions

continues to be the key driver in our overall business philosophy. As the technical
world within a company’s infrastructure continues to expand and become more

1
XYZ: Company and Client project details have been removed to protect confidentiality.
Specific references can be provided upon request.
1
CONFIDENTIAL Page 4 5/10/2010

complex, the company will continue to strive towards creating systems that will
make them more manageable and cost efficient for the future.

Proposal for the development of Automated Troubleshooting

solutions – an Expert System Approach
_________________________________________________________________
Overview:

As businesses become increasingly web-centric, multi-tiered and distributed, there

are increasing numbers of second-party software packages ‘strung’ together within
a single System. In many cases such as Internet Retail, the software modules are
interfaced with shopfloor, warehouse or shipping/ transportation equipment. Where
previously, debugging a problem could in most cases be confined to a single
platform, software application, or functional area, it is now necessary to analyze
and diagnose multiple sub-systems, platforms and hardware, and their interfaces.

Thus the level of Troubleshooting and routine Maintenance analysis has risen
dramatically from basic application debugging to include higher Operational levels.
People involved in maintenance and production support have to learn and
remember many routine tasks and emergency procedures. Support personnel have
to undergo exhaustive and expensive training in using repetitive fault- tracking
procedures to accommodate the increased number of sub-system components.

The knowledge transfer from experienced Support personnel to new hires and the
training involved in acclimatizing them to repetitive procedures is a costly exercise
in time and resources. It is almost entirely dependent on Human availability and
learning curves.

Some companies use hierarchical Support structures to deal with problem

resolution. Call Centers deal with user interface issues, and higher levels of
Support called in as needed. High importance is given to methodical debugging of
workflow. In many cases, event logs are also entered into a Knowledge Base for
future reference. Debugging tools exist at the technical ‘compiler’ level, and
Operating/ Troubleshooting manuals are available for investigating the process
flow.

Although the above basic methods are being used to alleviate the overload on
support personnel, there is a continuing dearth of intelligent enterprise-level
Decision Support tools in the field of Troubleshooting, Diagnostics and
Maintenance. There are very few digital diagnostic aids available that combine

Page 4 of 11 Printed: 05/10/10

CONFIDENTIAL Page 5 5/10/2010

advanced back-end technology with a user-friendly, intuitive front-end interface

suitable for use in mainstream Internal Technical Support.

In recent years, IT re engineering efforts have made rapid progress towards Object
Oriented paradigms, migration to "thin" clients, network connectivity, web- based
and wireless architecture. There are mainstream tools available for creating and
analyzing modular designs, and these support the thought process in these areas.
But in the areas of maintenance, diagnosis and problem resolution, there are few
resources readily available to match the growing levels of complexity in fault-
tracking procedures.

The unique advantage offered by Diagnostic Expert Systems could be used to

significantly improve troubleshooting in the interlinked & manifold systems of the
21st century. An Expert System can be used to diagnose all system parts
simultaneously, thus reducing the manual labor and time involved.

Similar to the use of Medical Diagnostic software by Doctors, Maintenance

checking and fault tracking systems used by Aircraft mechanics, or Artificial
Intelligence software embedded in applications like CRM, an AI Tool to perform
debugging and routine maintenance procedures would add great value in replacing
or supplementing manual tasks. It would encompass and address problem
resolution both at lower Technical and higher Operational levels.

Although the use of Artificial Intelligence-based systems in industry for Reliability

Centered Maintenance and diagnosis of machine parts is not new, and expert
systems have been used for highly specialized functions like maintenance of
Aircraft engines, their use as diagnostic tools in the mainstream corporate
environment is relatively limited.

And yet resolving issues in companies today can be as resource- consuming as

checking an Aircraft engine- on the positive side, more companies are adopting
systematic and structured approaches to workflow analysis albeit at a conceptual
level, thus lending their environments more easily to being modeled and diagnosed
similar to Aircraft parts.

In addition to the software, there is a critical need for rules, models and
standardized procedures based on which to automate troubleshooting and
escalation procedures. The proposed architecture for web- based deployment of a
diagnostic Expert System would enable these rules to be introduced and
implemented throughout the company, providing a framework for progressive
automation in this direction.

Page 5 of 11 Printed: 05/10/10

CONFIDENTIAL Page 6 5/10/2010

Salient Features
____________________________________________________________________________

This application would enable a company to organize internal Support tasks into
manageable, fixed and expanding Rules databases, plus use an AI engine as a time-
saving analytical aid, bringing the advancement of this underdeveloped area on par
with others like Billing, eFulFillment and eCRM. For companies that have already
initiated steps towards building Knowledge Databases of Rules and procedures,
this would be an efficient way of utilizing this data for routine & mission-critical
analysis.

The AI tool could also be used for generating executive-decision Reports from the
rules database for in-depth analysis.

Instead of depending on static documentation or memory, Support Personnel would

have recourse to an interactive Fault Tracking Tree that maps out a “Thought
Route” -from abstract to “nuts and bolts” levels. An important benefit of this
software is that it would streamline and automate not just Technical, but also
Administrative issues resolution, including Escalation and Emergency procedures
in the company.

When support personnel experienced in problem resolution leave a company, they

carry with them much of their Analytical and fault-tracking know-how – if these
are transferred into an Expert System at the time of resolution, new hires would
have this cumulative knowledge to rely on. This ongoing retention of all diagnostic
procedures in a logical, interactive and reusable format would greatly reduce
training time and and operating cost.

The operating plan to deploy this Expert System would involve creating a
knowledge base if there is none, then creating Fault Tree decision charts and
transferring these into the Expert System. Once in place, this system would imbibe
new problems and procedural rules as they arise.

Page 6 of 11 Printed: 05/10/10

CONFIDENTIAL Page 7 5/10/2010

System Design
____________________________________________________________________________

The basic components of this system as shown below, consist of a database of rules, a rules
engine to transfer rules into and from the database to a frontend, and a frontend interface that
can represent these rules in the form of Fault Tree decision charts.

FAULT TREE CHART

RULES
RULES BASE
ENGINE

USER PLACES A
USER CHANGES QUERY
A RULE

Page 7 of 11 Printed: 05/10/10

CONFIDENTIAL Page 8 5/10/2010

Market Analysis
____________________________________________________________________________

The increased dependence on IT capabilities within an organization has led to a

greater need for improved diagnostic and internal maintenance support. Growth is
expected to continue for the foreseeable future as the need for more efficient IT
departments continues to become a priority for many organizations.

As the marketplace continues to develop and evolve, our company will continue to
provide products and services that meet the needs of our customers, allowing them
to fully optimize their resources and contribute to a better managed enterprise.

Competition

We feel the key factors that will ultimately differentiate us from our competitors
are--

• improved methods provided to the customer to allow them to capture, store and
update their business rules extracted from an ongoing process, and the ability to
view diagnostic procedures in a graphic format through a user-friendly
interface.
• improved tools for creating Fault Tree charts for a diagnostic process, and
importing them into a knowledge base.
• improved modeling for routine maintenance and mapping of escalation
procedures within an organization.
• a flexible application that interfaces effectively with other enterprise systems,
and is designed to sustain and capture extensive changes in system architecture
in all functional areas.

With the explosive growth in application services, there is intense competition

among B2B and related companies to establish a superior presence. Our strengths
and capabilities lie in the groundbreaking solutions we provide in a critical niche
area of eBusiness, and the ability to provide not just a well-designed product but
also the complete service and support that the customers expect.

Page 8 of 11 Printed: 05/10/10

CONFIDENTIAL Page 9 5/10/2010

Target Markets

Buyers of the proposed Application can be classified into the following broad
categories:

• Call Centers.
• Later-stage startups with expanding, high-flux IT environments.
• Companies using IT systems based on a “chain” of disparate platforms and
software applications that need to be maintenance checked and analyzed.
• Examples of these types of systems can be found in Web- based Retail, Call
Center Support, Administrative and Manufacturing IT environments.
Operating Plan
_________________________________________________________________

The project implementation will be split into 6-month phases. Phase I will involve
the design and development of the AI Expert System as an Intranet application,
based on infrastructure provided by a chosen vendor company.

Phase II will focus on understanding the Client company's environment and

tailoring the Expert System application to fit it. It will begin with data gathering at
the client site and transferring data into rules for the knowledge bases, and charting
decision fault-tree charts based on ongoing processes, existing documentation and
event logs. Then the AI engine will be tested against these databases. By the end
of Phase II the client should have a fully functional tool that has been customized
and configured to fit their specific needs.
Sales Forecast

• We plan to sell this product at (TBD) 2per customer, which would include
streamlining their troubleshooting processes, and the deliverables spun off from
these comprising of knowledge bases, tools for creating and processing fault
tree decision charts .

• Based on the above Figures, we aim to build a customer base of atleast 4

customers during the (Marketing) Phase II of the first year.

2
TBD: Financial details have been removed from this document. They can be made available
upon request.

Page 9 of 11 Printed: 05/10/10

CONFIDENTIAL Page 10 5/10/2010

Projections
____________________________________________________________________________

Worst Case:

In the event that the product does not attract the anticipated number of customers
within the specified time, the basic technical infrastructure we would
have incorporated into the project could be easily re- cycled to serve
a wide variety of related functions that would benefit from the use of
AI- based software, like customer survey and profile Report
generation or network maintenance.

This project aims to deliver powerful results not just at a technical level in the form
of the product/ application itself, but also introduce the customer to
new methodologies and practices to automate their diagnostic and
escalation processes, and thus provide the basis for any future
development. Metaphorically, The Tools may die, but the
Methodology lives on!

Best Case:

Within the first year, the product will make an entry and establish itself as a
pioneer and distinctive presence in the marketplace for Artificial
Intelligence-based solutions for Troubleshooting IT Operations,
Systems Diagnostics and Maintenance.

Investment Offering:

Incubator assistance and Angel investment funding in the range of (TBD2) is sought
as startup and marketing capital for the first two phases, to be
completed within a six-month to one-year time period. Subject to the
first phase performance and growth factors (measured by number of
clients attracted) being positive and profitable, we aim to grow and
evolve as providers of customized Artificial Intelligence software
designed to fill a critical niche in the eBusiness environment.

2
TBD: Financial details have been removed from this document. They can be made available
upon request.

Page 10 of 11 Printed: 05/10/10